Tomographic electroencephalography/magnetoencephalography. Dynamics of human neural network switching.

New developments in multimodal registration of electroencephalography (EEG), magnetic resonance imaging (MRI), and positron emission tomography (PET) are presented as a method to create a tomographic EEG. Three-dimensional information about the x,y,z location of the sources of event-related potentials is corroborated through the use of experimental design and coregistration with MRI and PET. Once the three-dimensional location of event-related potential dipole sources is identified and corroborated, pseudoinverse procedures are used to derive a new EEG voltage sequence from each of the dipoles. Each derived EEG dipole time series is analogous to recording EEG from a deeply implanted electrode and constitutes a four-dimensional tomographic EEG (i.e., three-dimensional space plus time). EEG coherence and phase analyses are then performed on the dipole-derived time series to study the temporal and spatial dynamics of neural network switching during voluntary finger movements. The purpose of this article is to demonstrate a new method to exploit the time domain dynamics of neural network switching in behaving human subjects.